[Usability]Insert/overwrite, how should i do a usability study?

[Alan Horkan <horkana maths tcd ie>]

I would very much like help and advice on how to setup a usability study,
as I am having difficulty convince people of my concerns.

http://bugzilla.abisource.com/show_bug.cgi?id=3641
(dammit, Dom closed it again already!)

I have observed many many times users who have accidentally gone from
Insert mode (INS) to Overwrite mode (OVR) by accidentally hitting the
insert key.  It seems to me like a question that everyone asks at least
once.  It is quite hard for a user to identify and diagnose the problem
on their own.

If a user has been randomly hitting lots of keys on the keyboard to see
what happens they can by trial and error discover which key they need to
use to reverse the situation.

If a user has in the course of their work accidentally hit the insert key
they will not realize how they have gotten into this situation.

It is possible [1] but unlikely that users will guess the terms
Insert/Overwrite but even so, it is difficult to find the necessary
documentation (assuming it exists).

Better documentation would ease this situation, but it would be even
better if users did not have to look up the documentation (I am trying to
reach for a quote from Designing Web Usability by Jakob Nielsen, i read
something about this only this morning).

I don't think changing from INS to OVR is appropriate without more feedback
to the user.  The INS/OVR message is the status bar is cryptic and subtle
(in general Acronyms and Abbreviations are Evil), it may go unnoticed and
user might not connect the

For a usability study i could get users to turn away and put the computer
into overwrite mode and see how they solve the problem, but this seems
almost to crude and almost too easy.
I know i would need to record the study, if i have a it well prepared pen
and paper observations might be enough.  if there was some way to record
what was happening on screen would be good.

Perhaps i should be making this suggestion to the Sun people working on
OpenOffice rather than Abiword?

I would like to do this properly and scientifically even if that means not
getting the results i want.  Advice welcome.

Sincerely

Alan Horkan
http://advogato.org/person/AlanHorkan/

[1] http://abisource.com/mailinglists/abiword-user/2003/Jan/0064.html

Abiword, Microsoft Word, StarOffice Writer, OpenOffice.org Writer, Corel
WordPerfect, Kwrite and just about every other word processor are all
similarly unhelpful in this situation.